Automatic underwater vehicle landing on deep-ocean floor by successive trajectory learning control

For precision vehicle landing operation on the deep-ocean seafloor under uncertainties in current vector in water column in and estimated hydrodynamic drag, an automatic control technique, successive trajectory learning control (STLC) is proposed. The control enables the vehicle to descend along a planned trajectory and land automatically to the predetermined target point with prescribed accuracy. Convergence condition of STLC as well as design procedure for a real system are presented. A neutrally buoyant vehicle is controlled simultaneously with 6 thrusters. To confirm feasibility of the STLC in practice, a rigid cubic structure model was tested in a water basin subject to unknown current vector. Ultrasonic ranging system was arranged as LBL, and propeller-type thrusters were developed for tracking. The model successfully tracked the given trajectory and landed to its prescribed target on the basin floor, on its own STLC. Maximum speed of unknown steady current vector was 0.07 m/s for the test.